How to Run a Monthly Close That Takes an Hour

Why this matters

A monthly close that takes two evenings does not get done twelve times a year. It gets done seven or eight times, always late, always after the month it would have told you something about. By then the pricing decision it should have informed is already made and the error it would have caught is already repeated across four weeks of jobs.

The close is not long because closing is hard. It is long because most of it is not closing at all, it is searching: for a receipt, for whether a job got invoiced, for what a transaction was. Every one of those searches is a piece of work that should have happened weeks earlier, at the moment the information was free.

The rule this whole procedure rests on

A close is an assembly, not an investigation. If you are discovering things during the close, the close will take as long as the discovery does, which is unbounded. If everything was recorded correctly on the day it happened, the close is reading, comparing, and signing, and that fits in an hour.

So the steps below are not ordered by when you do them. They are ordered by what you lose if you skip them, roughly most expensive first, because the whole trick is knowing which upstream habit is buying you the most time. Two of them compete for the top spot depending on your shop: unbilled completed work is the largest single loss in a shop running dozens of tickets a week, while a broken cutoff costs more in a shop with long jobs that straddle month ends. Read the two lead-ins and decide which is yours. Five of the six are habits you hold during the month rather than tasks you do on close day.

1. Code at the point of spend (skipping this costs the most)

Every purchase gets its job number, or its overhead category, attached at the counter. Photograph the receipt against the transaction the same day, and if it belongs to a job, write the job number on it before you put your phone away.

What you lose by skipping it: the single biggest block of close time in every shop that has a long close. Three weeks later nobody can tell whether a fitting run belonged to the Tuesday service call or the Thursday install, so somebody guesses. That guess is not a bookkeeping detail, it is your job costing, which is your pricing model. The cost of skipping is not just the hours of chasing, it is that the answer you eventually record is fabricated.

How you know it is working: the count of transactions with no job number and no category at month end. Track it. It should trend toward single digits and stay there.

2. Reconcile weekly, not monthly

Match the bank and card feeds against the books once a week, in a short standing block. Reconciling means confirming that every transaction the bank shows exists in your books at the same amount and date, and that nothing in your books is missing from the bank.

What you lose by skipping it: not the time, which is roughly the same either way, but the ability to find anything. A duplicated charge found within a week is one phone call while the vendor still has it in front of them. Found at close, it is buried in four weeks of transactions and competing with everything else you are trying to finish. Worse, a reconciliation done once a month with a deadline attached tends to get forced: the difference gets plugged to an "ask my accountant" account and never revisited.

3. Keep a standing exceptions list

One running list of every transaction the person doing the coding could not place under your rule. Not a message, not a note in the accounting file, a list. The owner clears it on a fixed cadence.

What you lose by skipping it: the ambiguity does not disappear, it gets resolved silently by whoever is at the keyboard, consistently, in the direction that looks safest. Silent resolution is worse than a wrong answer because there is nothing to review. A list makes the ambiguity visible and, over two or three months, tells you exactly which sentence of your coding rule is unclear.

4. Invoice on completion, not in a batch

The invoice goes out when the job closes out, from the field or from the office the same day. The close then has nothing to reconstruct.

What you lose by skipping it: at close you have to walk every completed job and ask whether it was billed. That is a real reconstruction job in a shop running dozens of tickets a week, and the thing it finds late, unbilled work, is the most expensive thing in the whole procedure. Money you earned and never asked for does not show up as a loss anywhere. It simply is not there.

5. Keep owner money on its own card

One card, or one account, that only the business uses, and draws taken deliberately rather than by swiping whatever is in your pocket.

What you lose by skipping it: every mixed transaction becomes a question only you can answer, which means the close stalls waiting on you, and the ones you answer carelessly land in an expense account. A draw booked as an expense understates profit and overstates operating cost, so the shop looks worse than it is at exactly the moment you are trying to decide whether it can afford a hire.

6. Fix the close date and hold the cutoff

Pick a working day, the third or the fifth, and hold it. The cutoff rule matters more than the date: after the close is signed, nothing new gets dated into that month. A late receipt lands in the current month with a note.

What you lose by skipping it: a month that keeps moving. If entries continue to land in a closed month, every comparison you make against it is against a number that changed after you read it, and you will eventually make a decision on a figure that no longer exists in the file.

The hour itself

Once those six are in place, the close is a fixed sequence with a clock on it. This is the owner's hour, not the bookkeeper's. The bookkeeper's assembly work has already happened.

• 5 minutes. Confirm the cutoff held. Nothing new dated into the prior closed month.
• 10 minutes. Clear the exceptions list. Each item gets a decision and, if it repeats, a rule change.
• 15 minutes. Read the profit and loss statement as percentages against the trailing three months, not against last month alone.
• 10 minutes. Read the receivables aging. Name the three oldest balances and decide what happens to each.
• 10 minutes. Read the cash position and what is committed over the next four weeks.
• 10 minutes. Write down one question for your bookkeeper or accountant and one action you will take this month.

If the hour runs over, the overrun tells you which upstream habit failed. Time lost adjudicating exceptions means the coding rule is unclear. Time lost on the aging means invoices are going out late.

Worked example: where the nine and a half hours went

A six-person shop times its close honestly for one month before changing anything. Bookkeeper time is 7.0 hours: 2.5 chasing missing receipts and their job numbers, 2.0 matching completed jobs against invoices sent, 1.5 coding ambiguous transactions, 1.0 reconciling and producing statements. Owner time is 2.5 hours: 1.5 answering coding questions by text over four days, 1.0 actually reading anything. Combined, 9.5 hours.

Note what that split says. Of the owner's 2.5 hours, 1.5 went to supplying information nobody else had, and 1.0 to the only part that was ever the point.

They put the six habits in place. Receipts get coded at the counter. Reconciliation moves to a weekly 25-minute block. An exceptions list starts. Invoices go out at closeout.

The following month the close runs 1.5 hours of bookkeeper time and 1.0 hour of owner time, 2.5 hours combined. The weekly work added 4 sessions at 25 minutes, which is 100 minutes, or about 1.7 hours a month. Total monthly effort is 2.5 plus 1.7, about 4.2 hours against the original 9.5. That is about 44% of the previous total, and the 5.3 hours saved is the smaller half of the benefit.

The larger half: the statements now exist on the fifth working day instead of somewhere in the following month, and the guessing is gone. In the first close under the new routine the exceptions list ran 11 items, of which 7 were the same category of purchase. That is not 11 problems, it is one unclear sentence in the coding rule, and rewriting it dropped the next month's list to 4 items. Under the old routine those 11 would have been resolved silently and the pattern would never have surfaced.

Running it with nobody to hand off to

Plenty of shops have no bookkeeper, and the procedure still holds, but one thing changes and it is worth naming. When two people run the close, the exceptions list exists so that ambiguity crosses a desk. When you are both people, nothing crosses a desk, and the list has to do a different job: it is the record that you noticed you were guessing.

So keep it, and change how you use it. Write the item down the moment you cannot place it, do not decide it on the spot, and clear the whole list in one sitting at the close. Deciding 11 ambiguous items in one pass surfaces the pattern that deciding them one at a time over four weeks never will, because in a single sitting you can see that 7 of them are the same kind of purchase.

The second change is the review. A solo close has no second reader, so the trailing comparison is the only thing standing between you and a story you tell yourself. Read the current month against the three before it in percentages, every time, before you read the totals. The totals will always support whichever mood you brought.

When an hour is the wrong target

If you carry inventory of any real size, or run work in progress across month boundaries. Counting and valuing partially complete work is genuine close work, not a symptom of bad habits, and it does not compress. Budget for it separately rather than pretending the hour covers it.

In the first two or three months after any structural change. A new chart of accounts, a new field-service system, a change in how revenue books, or a new bookkeeper all mean the current month is not comparable to the last one, and the comparison is most of what the hour buys. Expect two to three hours until a clean trailing period exists again.

If you are behind. Do not try to close a month while three prior months are open. Catching up is a separate project with its own hours, and running it inside a monthly routine is how shops stay behind for a year.

References

• Generally Accepted Accounting Principles (GAAP), accounting period and cutoff concepts
• See related: Month-End Close Checklist for a Service Business; How to Close the Books at the End of the Week; When the Bank Balance Lies: Reconcile Discipline